The BL410 serves as an edge video analytics gateway, handling real-time processing of frontend video data to reduce reliance on cloud infrastructure, minimizing latency and bandwidth usage.
Case Details
The BL410 serves as an edge video analytics gateway, handling real-time processing of frontend video data to reduce reliance on cloud infrastructure, minimizing latency and bandwidth usage. Equipped with a 1 TOPS NPU and a Cortex-A55 quad-core processor, it efficiently runs common deep learning models (e.g., TensorFlow, PyTorch), supporting edge intelligence scenarios such as pedestrian detection, traffic flow counting, and security monitoring.
System Architecture
- Data Acquisition Supports multiple camera inputs, including IPC, USB Camera, and HDMI, catering to diverse video capture needs.
- BL410 AI Edge Gateway
- Utilizes NPU for accelerated target detection and behavior recognition in video streams.
- Stores key video frames or event clips locally (8/16/32GB eMMC).
- Deploys inference services via Docker or Node-Red, supporting MQTT and OPC UA protocols for seamless interaction with upper-level systems or cloud platforms.
- Data Upload and Management
- Uploads analysis results (e.g., alerts, statistical data) via 4G, WiFi, or Ethernet.
- Integrates with SCADA systems and major cloud platforms (e.g., AWS IoT, ThingsBoard, Ignition, Alibaba Cloud, Huawei IoT).
- Enables remote access and maintenance through the BLRAT tool.
Key Features
- Video Analytics: Supports 1080P@60fps decoding and 4K@60fps encoding, efficiently running detection and recognition algorithms.
- Local Preprocessing: Performs alert filtering, event classification, and keyframe extraction to optimize data processing.
- Data Integration: Interfaces with local alarms, access control, or lighting systems via RS485, CAN, DI/DO.
- Industrial-Grade Stability: Passes high/low-temperature, vibration, and surge tests, suitable for outdoor and industrial environments.
Application Scenarios
- Smart Security: Perimeter intrusion detection, video-linked alarms.
- Smart Transportation: Traffic flow monitoring, violation detection, license plate recognition preprocessing.
- Smart Factory: Production line video quality inspection, AI defect detection.
- Smart Campus: People flow counting, hazardous area monitoring.
